A business processing method, system, device and medium

Abstract

The invention discloses a business processing method, comprising the following steps: determining a first quantity parameter of acquiring a slice each time and a second quantity parameter of an object acquired each time in each slice; assigning a first weight to the first quantity parameter Obtain the third parameter of the number of fragments; enumerate the fragments corresponding to the third quantity parameter from the underlying multiple fragments, and enumerate the objects corresponding to the second quantity parameter from each of the enumerated fragments, and then get each A slice is a column, and the multiple objects corresponding to each slice are a matrix of rows; multiple objects corresponding to each row of the matrix are sequentially extracted; multiple objects extracted each time are processed according to the preset number of concurrent processing . The invention also discloses a system, computer equipment and readable storage medium. The solution proposed by the present invention is based on the idea of ​​enumerating a part of objects in a part of fragments at a time, and then enumerating them in turn, reducing unnecessary IO operations and reducing the impact of enumeration on performance.

Description

Technical field [0001] The present invention relates to the field of business treatment, and more particularly to a business processing method, a system, a device, and a storage medium. Background technique [0002] Distributed Object Storage System is a specific implementation of software definition storage. It is a new network storage. Based on its data that uses Key-Value to access, it is usually possible to store large-scale non-structural data (a bucket storage. The amount of data can even reach a hundred billion). For large-scale data, there is a means of automating it, which is generally called LifeyCyCle (Lifecycle Management). [0003] Key-Value's data access method guarantees that there is no additional index overhead that carries any key access value because of the expansion of the data amount, because the access to its data is not passed by the data to locate the data corresponding to the KEY. In which position is, it is positioned by a hash method. However, this does...

AI Technical Summary

Problems solved by technology

The life cycle currently adopts this sorting enumeration method. In addition to the above problems, there is another enumeration method. In a multi-version scenario, 1000 objects listed may be concentrated on one shard. At this time, if the life cycle traverses The subsequent processing of the object is concurrent, and the concurrent business of the life cycle is concentrated on one shard, which is completely inefficient. This is why it is not used to enumerate one shard before enumerating the next shard.

Images